Neuroblastoma is the most common extracranial solid tumor of childhood. Although patients with different expected survivals generally can be identified by clinical staging at diagnosis, individual stages include patients with quite different outcomes. Inter- and intra-stage diversity provides opportunities for identifying molecular genetic and biologic properties that are associated with distinct clinical phenotypes. such correlations can identify risk groups that otherwise are not recognizable, which can aid interpretation of clinical studies and guide development of new therapies. The long-term goal of the proposed studies is to develop tests that improve definition of risk groups so that the most appropriate and effective therapy can be given to each patient. The hypothesis of this proposal is that subsets of neuroblastomas can be identified by evaluating 1) genes that are critically involved in neuroblastoma cell growth, differentiation, and survival and 2) the ability of tumor cells to grow continuously in vitro. We plan to build upon our previous studies of the N-myc proto-oncogene, of the gp140trk high affinity receptor for nerve growth factor (NGF), and of tumor cell growth in vitro, which demonstrate the importance of these markers in prognostication. New data suggests that the NGF related neurotrophin, brain derived neurotrophic factor (BDNF) also may be a useful marker. Our specific aims are 1) to determine if the tumor phenotype defined by N-myc gene amplification, gp140trk expression, and continuous tumor cell growth in vitro correlates with disease progression during or after therapy; 2) to develop multivariate statistical models for predicting outcome based upon clinical and laboratory data; and 3) to determine in an exploratory study if BDNF expression defines risk groups and then to include this marker in the prospective large-scale study if appropriate. The Childrens Cancer Group performs phase III studies in which newly diagnosed patients receive defined therapy according to risk classification, which currently is based upon clinical stage, age, histopathology, serum ferritin, and N-myc gene amplification. Approximately 200 patients are registered annually in studies for low, intermediate, and high-risk neuroblastoma. Tumors and bone marrows have been obtained at diagnosis for a number of years, and these specimens continue to be obtained prospectively. Tumors will be tested to determine their phenotype with regard to N-myc amplification, gp140trk RNA expression, BDNF RNA expression, and growth in vitro. Bone marrows containing tumor cells will be used for assessing N-myc amplification (if necessary) and tumor cell growth in vitro. To determine if laboratory tests identify small, clinically important subsets, clinical data will guide selection of specimens for appropriate testing (case control analysis). for prospective, large-scale studies, laboratory and clinical data will be obtained independently and periodic interim analyses will be performed to determine if patients who fail therapy can be identified. We anticipate that these studies will improve prognostication. They also are likely to contribute to the development of novel and possibly more effective therapies for high risk patients.